If I asked a straight-forward question about your opinion on a simple topic, you would likely be able to articulate you thoughts. Now what if I asked that question, but before you gave an answer, you read the opinion of a stranger who answered the same question. Assuming you already have an opinion, you probably would not be affected by this, right? I have my doubts – perhaps the presentation of a relatively obvious argument might still influence someone’s opinion. Instead of sitting around wondering, I decided to find this out for myself, so I conducted a simple experiment.
Opinions can often change when presented with new novel information (“Oh, I never thought of that, you’re right!”) but we don’t think about how opinions may change even when presented with obvious information. I wanted to show that even when they hear an argument that they surely must have considered already, it may have an effect on how they view something.
I surveyed hundreds of students at an elite high school in the Kantou area, giving a single open-ended question. The median age was 16, and the range was 15-17.
All students were given the same question. “Should students in Japan be going to cram school?” Cram school, or juku in Japanese, is the system of organized after-school classes. Students were given up to 30 minutes to consider and write their answers. If more time was needed, they were allowed to submit their surveys later, but this rarely occurred.
After the question everyone received, students were presented with an opinion that either gave a strong argument for going to cram school (group A), a weak argument against going (group B), or a strong argument against going (group C). Then, each student was asked to use the sample opinion by either agreeing or disagreeing, and elaborating. It was made clear that they did not have to agree or disagree with the entire answer, and indeed some students gave reasonable answers about how the argument is not so black-and-white. However, the students who were not clearly either more for or against were not included in the data. The examples below show exactly what was asked for each group.
There were three groups in total. The total number of participants in Group A was 107, Group B was 35, and Group C was 105. In total, 247 students were entered in the data.
Group A is essentially the control group, because Japanese students from elite schools are likely to go to cram school. Or, at the very least, they are very likely to consider it as an option, and would therefore be more likely to agree with an opinion that students in Japan should go there. Note that the survey question referred to students in general – not necessarily the individual participating in the survey.
Despite group B having fewer participants, the students (rather, the classrooms) were randomly assigned to groups. It was decided that group B would not be dropped from the study, but I explain its inclusion below.
Here are the group differences:
Group A (Strong argument for)
Should students in Japan be going to cram school?
Mrs. Anonymous says “Students should go to cram school, because they need to be at a high academic level for university, or keep up with other students in this competitive job market.”
Group B (Weak argument against)
Should students in Japan be going to cram school?
Mrs. Anonymous says “Students should not be going to cram school, because it just adds more pressure, and it is simply unnecessary for them.”
Group C (Strong argument against)
Should students in Japan be going to cram school?
Mrs. Anonymous says “Students should not be going to cram school, because they can learn everything they need in regular school.”
I did not include participants in the data whose responses were unclear or ambiguous. This was sometimes a simple judgment call. For example, several did not explicitly say that they agree or disagree, but were deemed for or against, based on their arguments.
Group B Complications
The fact that there are two groups against cram schools (B & C) was not initially intentional. In one classroom, what is now Group B, there was potentially a problem in the way the opinion was interpreted by the participants. So I realized after they were asked complete their surveys – but before their data was collected – that the opinion given was considerably weak, and it was early enough to add a Group C with no negative consequences.
The reason the Group B opinion is weak is because it draws attention towards two points that can be quite easily refuted. That is, cram school does not just add pressure – it serves an important social role, such as for making friends with students from other institutions – and it is not simply unnecessary, since many students have anecdotes of benefiting from it greatly.
Having only groups A and B would be a problem because if students agreed with a strong argument for (group A) vs. only a weak argument against (group B), the results I expected to find would not reflect a tendency to agree so much as simply stuedents being persuaded by a better argument. Therefore, a strong argument against (group C) was subsequently used.
Group A consisted of 107 students. The breakdown of the results was among three classrooms in the same grade: Classroom 1 had 23 for and 16 against, classroom 2 had 18 for and 11 against, and classroom 3 had 26 for and 12 against.
In total, that’s 68 for, and 39 against. Graph I shows that almost two thirds of students thought that students should go to cram schools.
Group B consisted of 35 students. Their results were 13 for (37%), and 22 against (63%). This means that more people disagreed with the opinion that students should not go to cram schools. In other words, the students in Group B yielded virtually the same results as Group A – almost 2/3 think that Japanese students should go to cram schools.
Group C consisted of 105 students. The breakdown of the results was among three classrooms in the same grade: Classroom 1 had 16 for and 25 against, classroom 2 had 17 for and 10 against, and classroom 3 had 18 for and 19 against.
In total, that’s 51 for, and 54 against. Graph II shows that when a strong opinion is presented against cram schools, only half of the students were inclined to say that Japanese students should go to cram school.
If you combine groups B and C (which some may argue is useful) then the total is 64 for (46%), and 76 against (54%).
Just for the record, here is a sample response from a student from Group C, classroom 2:
Children have no time, for example, [to play] with their friends, reading books, doing club activities, and so on. Also, if they go to cram school, it costs a lot of money. They can learn everything they need in regular school. Cram school’s lessons finish at midnight. Children cannot go to bed early. So the next day, they cannot get up comfortably.
The hypothesis that Japanese students’ opinions would be influenced by the exposure of a stranger’s opinion was confirmed. This is interesting because most of us believe that we have a good rationale for our opinions. After all, we’re all reasonable people, aren’t we? Well, a simple one-line argument can evidently influence people, even when the participants are already more inclined to believe that Japanese students should go to cram school.
This isn’t all too surprising, because we have societal pressures to agree and resist confrontation, regardless of culture. As one summary of prominent social psychologist Robert Cialdini’s influence research said:
People love to say ‘yes’ to requests from people they know and like. And people tend to like others who appear to have similar opinions, personality traits, background, or lifestyle. More people will say ‘yes’ to you if they like you, and the more similar to them you appear to be, the more likely they are to like you.
All though we are influenced by those who are similar to us, the participants in my survey were more likely to agree (with the opinion against going to cram schools) despite having no affiliation with “Mrs. Anonymous” whatsoever. It was not a friend, or even a familiar name; and yet students agreed.
No study is ever perfect, and this has numerous limitations. For example, it would have been very beneficial to have included a weak argument for students going to cram school. That way, we would be more certain whether or not the strong argument is simply a very persuasive argument, or whether students just tend to be think that cram school is a good option.
Also, I had considered asking the students whether they themselves go to cram school or not. Some of them indicated that voluntarily, but it may have been interesting to see everyone’s answers. However, perhaps having them write that down might have produced more cognitive dissonance, whereby they would espouse the opinion that was most inline with their behavior. For example, if they were going to cram school, they would be more likely to say that students should go to scram school, and vice versa.
With that said, you might be wondering whether the 100+ participants in Group A just so happened to have more students attending cram school than those in Group C. So what if the students are all simply writing what they are doing? This may be going on to some extent, but the strength of this experiment is that the classrooms were randomly assigned to the groups, and there were a lot of participants.
Also, several students indicated that their opinion was actually the opposite of what they were doing (i.e., “I go to cram school but I probably shouldn’t, it costs too much money…” or “I don’t go to cram school, but I want to…”). So the best thing to confirm these results would be to replicate them.
It seems only natural that a culture as collectivistic as Japan would be persuaded to agree with someone (i.e., conform) despite being more likely to have the opposite inclination (hence the fact that so many Japanese students attend cram schools). But we can’t conclude that these results a) represent all of Japan, b) necessarily represent only Japan, or c) indicate that these students would have the same results on another opinion. We can only speculate. So I speculate that people, no matter what race or culture, prefer to agree more than disagree, with all things being equal.
In my study, I think the variables were not equal – I think more participants thought that students should go to cram school before the research began – which is why this study is important. That doesn’t necessarily mean that most of the students actually attend cram school, though. There are other factors in determining whether they actually go (e.g., time, money), but students in elite schools generally consider their academic prowess a priority worth the effort.
In the end, this information – if used properly – may be useful in helping you increase your level of persuasiveness with an argument in the future.
Don’t you agree?